| As the display center of the vehicle’s electronic control system,the quality of car instrument panel quality inspection plays an important role in safe driving.Under the general trend of the development of the Internet and electronic technology in the world,the display content of the instrument panel has also become more diversified,and the detection difficulty of the vehicle instrument panel has also increased.Traditional manual testing methods with low efficiency and low accuracy have been unable to cope with the increasingly complex testing problems.In this paper,in view of the problems of low accuracy and poor versatility of existing algorithms,the existing algorithms are researched and improved to improve the accuracy of detection and realize the automatic detection of automobile dashboards.The main research work is as follows:(1)According to the analysis of the needs of the detection system,a visual inspection system for the display status of the car dashboard is designed,the image acquisition equipment is selected,the communication between the car dashboard and the software system is established.And an experimental platform for visual inspection device for the display status of automobile dashboards is built.(2)The three types of display states of the pointer,signal light and nixie character of the car dashboard are detected by different image processing algorithms.In the automatic reading of the pointer,due to the low accuracy of the Zhang’s fast parallel thinning algorithm when thinning the straight line,an improved algorithm for straight line extraction is proposed,and the automatic reading of the pointer is realized,and the average relative error is reduced by 0.55%;A signal light image restoration method based on HSV image color space and improved OTSU method is used.HOG is used to extract the characteristics of the signal light to obtain the feature vector of each signal light.Combined with the SVM model,the signal light to be detected is identified and detected.The digital recognition method of BP neural network is used for identification and detection.(3)The visual inspection software system of automobile instrument panel is developed.The software interface is designed with MFC programming of VS2010,and the Open CV development environment is configured in the program,and the image processing algorithm and the software system are combined.Tests on different types of car dashboards.The experimental results show that the system accuracy in detecting car dashboards is better than manual detection methods and traditional detection algorithms,and the pass rate meets the expectations of the experiment,which realizes the automatic detection of car dashboards. |
PDF Full Text Request